1. Field of the Invention
The present invention relates to an orthogonal frequency division multiplexing (hereinafter abbreviated as “OFDM”) signal transmission system, and particularly to a layout relationship among antennas employed in an MIMO (Multiple Input Multiple Output) system using a transmitting device provided with a plurality of antennas and a receiving device provided with a plurality of antennas, and a transmission system using the same.
2. Description of the Related Art
FIG. 10 is a block diagram showing the relationship between a transmitting device and a receiving device employed in an MIMO system. The transmitting device 1000 comprises two antennas 1001 and 1002 and transmitters TX1 and TX2. Transmit data T1 is OFDM-modulated by the transmitter TX1, which in turn is transmitted through the antenna 1001 as a transmit signal W1. Further, transmit data T2 is OFDM-modulated by the transmitter TX2, which in turn is transmitted through the antenna 1002 as a transmit signal W2. On the other hand, the receiving device 1010 comprises two antennas 1011 and 1012, receivers RX1 and RX2 and an MIMO computing processor 1013. A signal received by the antenna 1011 is subjected to Fourier transformation or the like in the receiver RX1, which in turn is subjected to demodulation or the like in the MIMO computing processor 1013, so that transmit data is obtained. A signal received by the antenna 1012 is subjected to Fourier transformation or the like in the receiver RX2, which in turn is subjected to demodulation or the like in the MIMO computing processor 1013, so that transmit data is obtained. For the convenience of explanation in this case, the transmitting device 1000 and the receiving device 1010 will be explained assuming that they perform transmission/reception using the two antennas respectively.
Now consider where in FIG. 10, the function of transfer of data from the antenna 1001 to the antenna 1011 is assumed to be a, the function of transfer of data from the antenna 1002 to the antenna 1011 is assumed to be b, the function of transfer of data from the antenna 1001 to the antenna 1012 is assumed to be c, and the function of transfer of data from the antenna 1002 to the antenna 1012 is assumed to be d, respectively. The receiving device 1010 receives receiving signals via the plurality of receiving antennas 1011 and 1012 and extracts necessary signal components from the overlapped received signals, using an inverse matrix of the transfer functions for their propagation paths, respectively. A computing process for extracting the necessary signal components from the overlapped received signals will be explained below.
In a noise-free ideal state, the signals that overlap each other at the receiving antennas respectively satisfy the following equation:
                              (                                                                      R                  ⁢                                                                          ⁢                  1                                                                                                      R                  ⁢                                                                          ⁢                  2                                                              )                =                              (                                                            a                                                  b                                                                              c                                                  d                                                      )                    ⁢                      (                                                                                T                    ⁢                                                                                  ⁢                    1                                                                                                                    T                    ⁢                                                                                  ⁢                    2                                                                        )                                              (        1        )            
Then, the computing process for extracting the necessary signal components from each of the overlapped signals in the ideal state performs estimation of the transfer functions for radio wave propagation by means of the receiving device through the use of a pilot signal or the like. Next, the following arithmetic operation is performed using estimated values thereof:
                              (                                                                      T                  ⁢                                                                          ⁢                  1                                                                                                      T                  ⁢                                                                          ⁢                  2                                                              )                =                              1                                          a                *                d                            -                              b                *                c                                              ⁢                      (                                                            d                                                                      -                    b                                                                                                                    -                    c                                                                    a                                                      )                    ⁢                      (                                                                                R                    ⁢                                                                                  ⁢                    1                                                                                                                    R                    ⁢                                                                                  ⁢                    2                                                                        )                                              (        2        )            
Since noise components are contained in a signal to be demodulated upon actual communication, the equation (2) is represented as follows. In the following equation, n1 and n2 indicate noise components respectively.
                              (                                                                      T                  ⁢                                                                          ⁢                                      1                    ′                                                                                                                        T                  ⁢                                                                          ⁢                                      2                    ′                                                                                )                =                              1                                          a                *                d                            -                              b                *                c                                              ⁢                      (                                                            d                                                                      -                    b                                                                                                                    -                    c                                                                    a                                                      )                    ⁢                      (                                                                                                      R                      ⁢                                                                                          ⁢                      1                                        +                                          n                      ⁢                                                                                          ⁢                      1                                                                                                                                                              R                      ⁢                                                                                          ⁢                      2                                        +                                          n                      ⁢                                                                                          ⁢                      2                                                                                            )                                              (        3        )            
The following equations are derived from the equations (1) and (3):
                              T          ⁢                                          ⁢                      1            ′                          =                              T            ⁢                                                  ⁢            1                    +                                    1                                                a                  *                  d                                -                                  b                  *                  c                                                      ⁢                          (                                                d                  *                  n                  ⁢                                                                          ⁢                  1                                -                                  b                  *                  n                  ⁢                                                                          ⁢                  2                                            )                                                          (        4        )                                          T          ⁢                                          ⁢                      2            ′                          =                              T            ⁢                                                  ⁢            2                    +                                    1                                                a                  *                  d                                -                                  b                  *                  c                                                      ⁢                          (                                                                    -                    c                                    *                  n                  ⁢                                                                          ⁢                  1                                -                                  a                  *                  n                  ⁢                                                                          ⁢                  2                                            )                                                          (        5        )            
In the MIMO system as is understood from the above equations (4) and (5), when the noise is assumed to be a normal distribution, the more the denominator (a*d−b*c) decreases, the more the dependence on the noise components increases, and hence the receiving characteristics are degraded.
Incidentally, a patent document 1 (Japanese Unexamined Patent Publication No. 2002-374224) discloses that in an OFDM signal transmission system, an inverse matrix computation is performed using transfer functions to restore and demodulate a transmit signal.
Also a patent document 2 (Japanese Unexamined Patent Publication No. 2003-244056) discloses a base station which receives receiving signals using an antenna perpendicular to ground and an antenna parallel to ground in a wireless communication system.
However, both the patent documents do not have the aim of arranging antennas considering transfer functions for radio wave propagation in an MIMO system to thereby reduce noise components and avoid degradation of receiving characteristics.